investigations of hydrodynamics of fixed bed reactor

Investigations of Hydrodynamics of Fixed Bed Reactor: Counter-Current Flow

by

Norafheeza Norazahar

Dissertation submitted in partial fulfillment ofthe requirements for the

Bachelor ofEngineering (Hons)(Chemical Engineering)

JANUARY 2006

Universiti Teknologi PETRONASBandar Sen Iskandar

31750 Tronoh ^Perak Darul Ridzuan '' '. _

Dissertation . ECB 5034Final Year ResearchProject

Investigations of Hydrodynamics of Fixed Bed Reactor: Counter-Current Flow

by

Norafneeza Norazahar

Approved by,

A project dissertation submitted to the

Chemical Engineering Programme

Universiti Teknologi PETRONAS

in partial fulfillment of the requirement for the

BACHELOR OF ENGINEERING (Hons)

(CHEMICAL ENGINEERING)

rati Mellon

Project Supervisor

UNIVERSITI TEKNOLOGI PETRONAS

TRONOH, PERAK

January 2006

Dissertation ECB 5034 Final Year Research Project

This is to certify that I am responsible for the work submitted in this project, that the

original work is my own except as specified inthe references and acknowledgements,

and that the original work contained herein have not been undertaken or done by

unspecified sources or persons.

NORAFNEEZA NORAZAHAR

11

Dissertation . ECB 5034 Final Year ResearchProject

The residence time distribution (RTD) of a pilot plant scale packed bed reactor was

studied in the present study using air - water system to investigate the effect various

gas/liquid volumetric ratio on the residence time distribution. The ranges of air -

water ratio is chosen to simulate the hydrogen/oil ratio of typical bench scale hydro-

processing units, noting the difference that might arise due to the different physical

properties of theliquid. The experiments were conducted incounter-current mode and

the effects of gas/liquid velocities on the hydrodynamics of the reactor were

investigated. This study is done using the tracer method - specifically the pulse input

method. The packed bed reactor used in this study is packed with 8x8 mm Raschig

rings. Results show that the increasing gas/liquid ratio increases the mean residence

time. However, operating liquid hold-up and the value of Peclet number, which gives

an indication to the degree of axial dispersion in the system, increase. Second moment

analysis of the reactor, namely the variance of the system, shows a reduction in value

with decreasing gas/liquid ratio suggesting wider liquid distribution through the

system at lowgas flow rates. The discrepancies in experimental results suggested that

there are conditions to be altered in order to eliminate the inconsistency.

in

Dissertation . . ECB 5034 Final Year ResearchProject

TM

The author would like to be grateful to The Most Merciful, Allah S.W.T, for giving

the chanceto conduct and successfully completethis final year research project.

The author's outmost appreciation is for the project supervisor, Ms. Nurhayati

Mellon, who has given the trust and supervision to conduct such outstanding

experiment for this Final YearResearchProject.

To Universiti Teknologi PETRONAS, the author is thankful for all the equipment

provided in order to complete the experiments. Also to Reaction Engineering

laboratory technicians, who have helped in various ways in providing sufficient

assistance and supports.

To supportive colleagues and families, the author mostly appreciates the motivations

and advices.

Also to all individual involved directly or indirectly in this project, the author is

thankful for the contribution in progression and completion of this Final Year

Research Project.

IV


TABLE OF CONTENTS

CERTIFICATION OF APPROVAL - i

CERTIFICATION OF ORIGINALITY ii

ABSTRACT i«

ACKNOWLEDGEMENTS iv

LIST OF FIGURES AND TABLE ~ 1

ABBREVIATIONS AND NOMENCLATURES 3

CHAPTER 1: INTRODUCTION. 5

1.1 Background Study 5

1.2 Problem Statement..... 6

13 Significant of Project »7

1.4 Objectives 7

1.5 Project Scope 7

CHAPTER 2: LITERATURE REVIEW 9

2.1 Packed Bed 9

2.2 Packed Column Hydraulics 10

23 Operating Liquid Hold-up — 11

2.4 Residence Time Distribution (RTD)..., 12

2.5 Measurement of RTD .13

2.6 Characteristics of RTD 17

2.7 Axial Dispersion ». * 22

2.8 Stagnant Zone Volume 25

2.9 Late Mixing 28

2.10 Early Mixing 29

CHAPTER 3: PROJECT OVERVIEW - ....30

3.1 Research Project Work — —30

3.2 Experimental Set-up , 31


CHAPTER 4: RESULTS AND DISCUSSION 34

4.1 Effect of Gas/Liquid Ratio on Pressure Drop and

Operating Liquid Hold-up —. ....— 34

4.2 Effect of Gas/Liquid Ratio on Moments ofRTD 36

4.3 Effect of Gas/Liquid Ratio on Axial Dispersion and

Stagnant Zone Volume by RTD Analysis ...—.44

CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS 47

5.1 Conclusions.... 47

5.2 Recommendations .... . .—49

REFERENCES 51

APPENDICES 52

APPENDIX A RTD Analysis for Gas/Uquid Ratio of 150 . .—52

APPENDIX B RTD Analysis for Gas/Liquid Ratio of 200 54

APPENDIX C RTD Analysis for Gas/Liquid Ratio of 250 56

VI

Dissertation ECB 5034 Final Year ResearchProject

LIST OF FIGURES

Figure 2.1 Schematicofpacked column

Figure 2.2 RTD measurements

Figure 2.3 Typical RTDs

Figure2.4 Variance for matching theoretical curves

Figure 2.5 The spreading of tracer according to the dispersion model

Figure 2.6 Representation of the dispersion (dispersed plugflow) model

Figure Z 7Non-ideal flow patterns, which may exist in process equipment

Figure 3.1 Research methodology

Figure 3.2 Process Diagram for RTD Studied in Tubular Reactor (BP 112)

Figure4.1 Effect of gas/liquid ratio on pressure drop at constant liquid flow rate of

0.05 and 0.10 LPM

Figure 4.2 Effect of gas/liquid ratioon operating liquid hold-up at constant liquid

flow rates of0.05 and 0.10 LPM

Figure 4.3 E curve for gas/liquid ratio of 150

Figure 4.4 E curve for gas/liquid ratio of200


Figure4.6 Mean residence time for gas/liquid ratioof250

Figure 4.7 Effect of gas/liquid ratio on mean residence time of liquid at constant

liquid flow rates of 0.05 and 0.10 LPM

Figure 4.8 Variance curve for gas/liquid ratio of 250

Figure4.9 Effect of gas/liquidratio on variance at constantflow rates of0.05 and

0.10 LPM

Figure 4.10 Skewnesscurve for gas/liquid ratio of250

Figure 4,11 Effect ofgas/liquid ratio on skewness at constant liquid flow rates of 0.05

andO.lOLPM

Figure 4.12 Effect of gas/liquid on liquid Pecletnumber at constant liquid flow rates

of0.05and0.10LPM

Figure4.13 Effect of gas/liquid on liquidPeclet numberfrom other research


Figure 4.14 Effect of gas/liquid ratio on stagnant zonevolume at constant liquid flow

rates of0.05 and 0.10 LPM

FigureA-l Conductivity against time for gas/liquid ratio of 150

FigureA-2 Mean residence time curve for gas/liquid ratio of 150

FigureA-3 Variance curve for gas/liquid ratio of 150

FigureA~4 Skewness curve for gas/liquid ratio of 150

Figure B-l Conductivityagainst time for gas/liquid ratio of 200

FigureB-2 Mean residence time curve for gas/liquid ratio of 200

Figure B-3 Variance curve for gas/liquid ratio of200

Figure B-4 Skewnesscurve for gas/liquid ratio of200

Figure C-l Conductivity against time for gas/liquid ratio of 250

Table 3.1 Experiment gas and liquid flow rates with specified gas/liquid ratio


O F

LPM Liter per minute

LHSV Liquid Hour Space Velocity

dc Column diameter

dP Particle diameter

AP Pressure drop

Ha Operating liquid holdup

4 Height ofcolumn

S Fractional void volume

VS Absolute viscosity ofgas

Pi Absolute viscosity of liquid

Us Superficial gas velocity

u, Superficial liquid velocity

Dl Effective particle diameter

Gs Mass flow rate of gas

G, Mass flow rate of liquid

Pg Density of gas

Pi Density Of liquid

g Gravitational acceleration

Re Reynolds number

D Axial dispersion coefficient

L Distance between the tracer injection point and conductivity

measurement probe

u Mean real liquid velocity

tm Mean residence time

a2 Variance of the E curve

s3 Skewness of the E curve

Dissertation ECB 5034 Final Year Research Proiect

a2Q Variance oftheE curve for dimensionless time units

C Concentration

Ci Concentration at time t\

HRT Hydraulic retention time

Td Theoretical meanresidence time

Q Volumetric flow rate of liquid

V Volume of column


CHAPTER 1

INTRODUCTION

1.1 Background Study

Multiphase systems are very often applied in biotechnology (immobilized microbes or

enzymes on a solid support used as a catalyst for microbic or enzyme fermentations)

and chemical processes where the reactants andcatalyst are present in different phases

(gas-liquid and solid-catalyst). The reaction rate of such processes usually depends

on the specific contact area between the phases, such as fluid and solid, as well as on

the mass transfer rate from one phase to the other. External mixing in a reactor of

different and specific construction and usage could generate phase dispersion.

The flow of phases in continuous processes is usually not ideal and in some cases,

such deviation from the ideal flow pattern (plug flow or ideal mixing) might be

substantial. Such deviations from the ideal type of fluid flow are usually channeling,

recycle flow or stagnant zone formation inside the reactor or some other type of

equipment used in the chemical process industry (distillation columns, absorbers,

adsorbers, heat exchanger). When such non-ideal behavior of fluid flow occurs, an

important decrease ofprocess unit productivity may be detected.

Knowing the effect of non-ideal flow is essential in the case of scale-up design. In

many cases, it is enough to have information about the residence time distribution of a

fluid in a reactor determined by stimulus-response methods with the appropriate

tracer. The diminishing effects of non-ideal flow in the design of industrial units

might lead to enormous errors in their exploitation.

The laboratory or bench-scale units with lower capital investments are preferred for

these investigations. Studies in these units also involve lower operating cost, lesser

inventory ofchemicals, safer operations and reduced risk of hazards.


Packed bed reactors have found wide application in industrial practice and are

considered more efficient than the other types of contactors, because of increased

efficiency brought about by the use of packings. Packed bed reactors are vertical

towers equipped with suitable liquid and gas distributors and have grids for the

support of packing materials, which help to vary the local velocity of the continuous

phaseand restrictthe re-circulation and back-mixing in the phases considered.

A vast amount ofresearch has been carried out in three phase packed bed reactors. All

liquid residence time distribution (RTD) measurements were conducted at room

temperature and near atmospheric pressure. The tracer technique was chosen for

determination of the needed liquid flow parameters. The change in electrical

conductivity of a potassium chloride (KC1) tracer was simultaneously measured

downstream/upstream ofthe injection point at both the top and the bottom of the bed

using computer-monitored conductivity sensors. A brief tracer pulse was syringe-

injected into the liquid inlet pipe. The signals of the sensors were amplified and

transmitted to a computer by a data acquisition system (DAS). A residence time

distribution model to describe the liquid flow in a fixed bed reactor packed with

porous particles and operated under both partially and fully wetted conditions was

proposed based on a simple representation of the liquid flow structure. This model

assumes that the external liquid stream is divided into two zones: a dynamic zone in

which the liquid flows through the bed as a piston flow with axial dispersion and a

stagnant zone in direct contact with the dynamic zone and the solid particle. The

dynamic evolution ofthe tracer concentration in the particles was described in terms

of diffusion phenomena [1].

1.2 Problem Statement

Hydrodynamics play an important part on the performance ofpacked bed reactor. The

knowledge of the real flow pattern and the active volume of packed bed reactor are

critical information to scale-up and modeling purposes. The performance of packed

bed reactor relies on hydrodynamics, such as mixing, mass and heat transfer, other

than its microbial kinetics and physical phenomena. As for mixing, there exist two


categories, which are micromixing and macromixing. The effective method of

examining the macromixing behavior of reactor is Residence Time Distribution

(RTD) analysis.

1.3 Significant of Project

The RTD of a bench-scale multiphase reactor has been investigated mainly for

industrial scale-up purposes. The successful design of commercial reactors involved

generation of reliable data in laboratory-scale reactors and scaling up ofthese data for

larger units. The study of effects of gas and liquid flow rates on various

hydrodynamics parameters utilizing the RTD technique, using tracer is important for

the performance ofthe reactor. Future work may be based on the development of this

study.

1.4 Objectives

The objectives of investigations of hydrodynamic in a packed bed reactor of counter-

current flow are:

1. To investigate the effect of liquid and gas flow rate on pressure drop and

operating liquid hold-up.

2. To characterize residence time distributions (RTD) of reactor by three

moments pertaining mean residence time, variance, and skewness.

3. To perform the residence time study (tracer study) in order to analyse the

effect of gas/liquid volumetric ratio on the hydrodynamics of the reactor.

1.5 Project Scope

The scope ofresearch, which was successfully accomplished:

1. Understand the principle and characteristics of hydrodynamics.

2. Investigate the effect ofgas and liquid flow rates towards residence time

distribution (RTD) and further investigation on three moments ofRTD

Dissertation ECB 5034FinalYearResearch Proiect

3. Determine the liquid hold-up in a packed bedreactor at gas and liquid flow

rates.

4. Familiarize the concept ofback-mixing and stagnant zone volume.

5. Understand the concept of axial dispersion on gas/liquid ratio to observe the

mixing behavior of reactor.

6. Determine pressure drop in packed bed reactor at different gas and liquid flow

rates.


CHAPTER 2

LITERATURE REVIEW

2.1 Packed Bed

Packed bed columns for gas-liquid contacting are used extensively in chemical

process, especially in separation and reaaion. Usually the columns are filled with a

randomly oriented packing material, but for an increasing number of applications the

packing is very carefully positioned in the column. The packed column is

characteristically operated with counter flow ofthe phases.

The packed column is a simple device as shown in Figure 2.1. A typical column

consists of a cylindrical shell, containing a support plate for the packing material and

a liquid distributing device designed, to provide good liquid distribution on the

system. Devices may be added to the packed bed to provide redistribution of liquid

that might channel down the wall or otherwise become maldistributed. Several beds,

each with liquid distributor and support device, may be used within the same column

shell. For example, a distillation column with rectifying and stripping zones requires a

minimum oftwo beds.

Packed columns are usually specified when plate devices would not be feasible

because of undesirable fluid characteristic or some special design requirement.

Conditions favouring packed columns are:

• For columns less than 0.6-m (2.0-ft) diameter, packings are usually cheaper

than plates unless alloy-metal packings are required.

• Acids and many other corrosive materials can be handled in packed column

because construction can be of ceramic, carbon, or other resistant materials.

• Packings often exhibit desirable efficiency-pressure-drop characteristic for

critical vacuum distillations.


Liquids, tending to foam, may be handled more readily in packed columns

because ofthe relatively low degree of liquidagitation by the gas.

Hold-up of liquid can be quite low in packed columns, an advantage when the

liquid is thermally sensitive.

Support plate

Vapor Out

TLiquid Outlet

Distributor

Liquid Inlet

Packing

Vapor In

Figure 2.1 Schematicofpacked column

2.2 Packed Column Hydraulics

The pressure loss accompanying the flow ofgas through packed columns hasbeen the

subject of many theoretical analysis and experiment investigations to try to find a

suitable mathematical expression to predict the pressure drop caused by both kinetic

and viscous energy losses.

A very successfiil attempt is that of Ergun [2] which is included in the Perry's

Handbook. The Ergun equation is:

E S s3 Dl+ 1.75 (1)

10

Dissertation ECB 5034 Final Year ResearchProiect

The Ergun equation gave very good results in the whole range of Reynolds numbers

from 1 to 100,000. Also, it should be noted that the effectivediameter is equal to real

diameter onlywhenthe particles are spherical; for all other shapes the Dp is define as

Dp = 6Vp/Ap, where Vp is the volume of particles and Ap is the external surface of

particle. Ergun equation assumed equivalent pressure drop regardless of any type of

flow regimes.

Turpin and Huntington [3] also gave a single relation for pressure drop valid for all

the regimes, in terms of adimensionless parameter, Z=Re),167 /Re™ . On the other

hand, Varma et al. [4] developed an empirical equation for predicting the transition

from one flow regime to another. It presented typical variation of frictional pressure

drop with liquid and gas flow rates respectively for bubble flow, pulse flow and spray

flow. It is seen that though the pressuredrop increases with the gas and liquid rates in

all the regimes, its variation differs for the different flow regimes. For example, the

pressure drop increases rapidly withthe gas rate in the spray flow as compared to its

increase in pulse flow and in bubble flow.

However, it is noted that the transition between the flow regimes is not sharp and

occurred over a small range in gas and liquid flow rates. Thus, all flow regimes in

counter-current downflow can be assumed equivalent in this experiment, as in Ergun

[2] principle, which has been used widely in several researches.

In experiment, pressure drop is directly obtained from the differential pressure reading

at control panel or via the Data Acquisition System (DAS). The pressure drop

reading, in unit of mbar, is taken at time interval of one minute, and readings are

averaged for one value of pressure drop for every variation of gas and liquid flow

rates.

2.3 Operating Liquid Hold-up

The dynamic liquid holdup at different flow rates of the phases, refers to Iluita et al

[1], is the volume of the liquid drained after the inflow was stopped after reaching

steady state. The static liquid holdup was obtained from the difference between the

11


external void volume of the bed and that of the liquid drained after the reactor had

been totally flooded with liquid.

Experimentally, the operating liquid holdup of liquid is the portion of liquid that is

drained out of the catalyst bed when both gas and liquid flows are stopped. The

operating liquid holdup is an important parameter influencing the rate of reaction in a

gas-liquid-solid multiphase reactor. The operating liquid holdup of liquid isdefined as

the ratio of the volume of the free-drained water to the total volume of the packed

bed.

Chander et al. [5] determined the effect of liquid space velocityon holdup and proved

that the operating liquid holdup increased with liquid space velocity. Thus, higher

liquid flow rate could increase the reaction rate. Besides, the studies also showed that

liquid holdup for the upflow mode of operation was reduced when smaller size of

particles was used. Stiegel and Shah [6] also have reported the decrease of liquid

holdup with the decrease in particle size for the upflow mode of operation. It is also

observed that studies by Chander et al. [5] showed that when catalyst bed was diluted

with smaller size of particles, the effect of space velocity on operating liquid holdup

was very small or negligible.

Chander et al. [5] also studied the effect of gas/liquid ratio on operating liquid holdup,

which resulted that the liquid holdup decreased with increasing gas flow rate for the

upflow mode when the bed waspacked witha larger size of diluent.

2.4 Residence Time Distribution (RTI>)

The RTD of a reactor is a characteristic of the mixing that occurs in the chemical

reactor. There is no axial mixing in a plug-flow reactor (PFR), and this omission is

reflected in RTDwhich is exhibited by this classof reactors. The constant stirred type

reactor (CSTR) is thoroughly mixed and possesses a far different kind of RTD than

the plug-flow reactor. The RTD of a reactor yields distinctive clues to the type of

mixing occurring within it and is one of most informative characterizations of the

reactor.

12


2.5 Measurement of RTD

TheRTD is determined experimentally by injection of an inert chemical, molecule, or

atom, called a tracer, into the reactor at sometime, t= 0 and then measuring the tracer

concentration, C, in the effluent stream as a function of time. In addition to being a

non-reactive species that is easily detectable, the tracer should have physical

properties similar to those of the reacting mixture and be completely soluble in the

mixture. The latter requirements are needed so that the behavior of tracer will

honestly reflect that of the material flowing through the reactor. The two most used

methods of injection are pulse input and step input;

2.5.1 Pulse Input

In a pulse input, an amount of tracer N0 is suddenly injected in one shot into

the feedstream entering the reactor in as short a time as possible. The outlet

concentration is then measured as a function of time. Typical concentration-

time curves at the inlet and outlet of an arbitrary reactor are shown in Figure

2.2. The effluent concentration-time curve is referred to as the C curve in RTD

analysis. The injection of a tracerpulse shall be analysed for a single-input and

single-output system in which only flow (i.e. no dispersion) carries the tracer

material across system boundaries. First, an increment of time At is chose to

be sufficiently small that the concentration of tracer, C(t), exiting between

time / and t + At is essentially constant. The amount of tracer material, AW,

leaving the reactor between time / and t + At is then

AN =C(t)vAt (2)

where v is the effluent volumetric flow rate. In other words, AN is the

amount of material that has spent time between time t and t + At in the

reactor. If the term is divided by the total amount of material that was injected

into reactor, N0, then

W=VMA{ (3)

13


which represents the fraction of material that has a residence time in the

reactor between time / and t + At.

Feed x">s•JjfrnJ. Jp£ frn-n.w>a*[—i. jfrlh-j

r f

_Y

I (niceiinn 1l_l_ „J

Puise injection

¥

Cj{3j injection

0

neacror

Putse response

-if

*tep response

Figure 2.2 RTD measurements

For pulse injection, it is defined

JV.

So that,

AN =E(t)At

(4)

(5)

The quantity E(t) is called the residence-time distribution function. It is the

function that describes in a quantitative manner how muchtime different fluid

elements have spent in the reactor.

14


If N0 is not known directly, it can be obtained from the outlet concentration

measurements by summing up all the amounts of materials, AN, between time

equal to zero and infinity. Writing equation (2) in differential form yields,

dN =vC(t)dt (6)

And then integrating,

to

N0^\vC(t)dt (7)o

The volumetric flow rate is usually constant, so E(t) can be defined as

£('hc(t)

jC(t)dt(8)

The integral in the denominator is the area under the C curve.

An alternative way of interpreting the residence-time fianction is in its integral

form:

Fraction of material leaving the

reactor that has resided in the reactor

for time between /, and /:

JE(t)dt (9)

It is known that the fraction of all the material that has resided for a time t in

the reactor between t = 0 and / = qo is 1; therefore,

JE(t)dt =\ (10)

Theprincipal potential difficulties with the pulsetechnique lie in the problems

connected with obtaining a reasonable pulse at a reactor's entrance. The

injection must take place over a period, which is very short compared with

residence times in various segmentsofthe reactor or reactor system, and there

must be a negligible amount of dispersion between the point of injection and

the entrance to the reactor system. If these conditions can be fulfilled, this

technique represents a simple and direct wayof obtaining the RTD.

15


There are problems when the concentration-time curve has a long tail because

the analysis can be subject to large inaccuracies. This problem principally

affects the denominator of the right-hand side of equation (8), i.e. the

integration of the C(t) curve. It is desirable to extrapolate the tail and

analytically continue the calculation. The tail of the curve may sometimes be

approximated as an exponential decay. The inaccuracies introduced by this

assumption are very likely to be much less than those resulting from either

truncation or numerical imprecision in this region.

2.5.2 Step Tracer

The meaning of the RTD curve is previously discussed, now a more general

relationship between a time varying tracer injection and the corresponding

concentration in the effluent will be formulated. It should be stated without

development that the output concentration from a vessel is related to the input

concentration by the convolution integral.

C„M =ic,„(t-t')E(f)df (11)ctrCc

fadO

The inlet concentration most often takes the form of either perfect pulse input •»

(Dirac delta function), imperfect pulse injection, refer to Figure 2J, or step

input.

Step input in the tracer concentration will be analysed for a system with a

constant volumetric flow rate. Consider a constant rate of tracer addition to a

feed which is initiated at time / = 0. Before this time, no tracer was added to

the feed. Stated symbolically,

Co(f)=0 /<o

Ce(/) =constant ?>0

The concentration of the tracer in the feed to the reactor is kept at this level

until the concentration in the effluent is indistinguishable from that in the feed;

16

Dissertation __ ECB 5034Final Year ResearchProiect

the test may then be discontinued. A typical outlet concentration curve for this

type of input is shown in Figure 2.3.

Becausethe inlet concentration is constant with time, CQ, the integral is taken

outside the integral sign, i.e.,

Cout=C.]E{t')dt>

Dividing by CQ

r

This expression is differentiated to obtain the RTD function E(t):

(13)

=\E{t')dt' =Fit) (12)step 0

m-rrstep

The positive step is usually easier to carry out experimentally that the pulse

test, and it has the additional advantage that the total amount of tracer in the

feed over the period of the test does not have to be known as it does in the

pulse test. One possible drawback in this technique is that it is sometimes

difficult to maintain a constant tracer concentration in the feed. Obtaining the

RTD from this test also involves differentiation of the data and presents an

additional and probably more serious drawback to the technique, because

differentiation of data, on occasion, can lead to large errors. A third problem

lies with the large amount of tracer required for this test. If the tracer is very

expensive, a pulsetest is usuallyused to minimize the cost.

2.6 Characteristics of RTD

Sometimes E{t) is called the exit-age distribution function. Ifthe 'age' ofan atom is

regarded as the time it has resided in the reaction environment, the E(t) concerns the

age distribution of the effluent stream. It is the most used of the distribution functions

connected with reactor analysis because it characterizes the lengths of time various

atoms spend at reaction conditions.

17


Figure 2.3 illustrates typical RTDs resulting from different reactor situations. Figure

2.3 (a) and (b) correspond to nearly ideal PFRs and CSTRs respectively. In Figure

2.3 (c), it is observed that a principal peak occurs at a time smaller than the space-

time, r = V/v (i.e. early exit of fluid) and also that fluid exits at a time greater than

space timer. This curve is representative of the RTD for a packed-bed reactor with

channeling and dead zones. One scenario by which this situation might occur is

shown in Figure 2.3 (d). Figure 2.3 (e) shows the RTD for the CSTR in Figure Z3

ff) which has dead zones and bypassing. The dead zone serves to reduce the effective

reactor volume indicating that the active reactorvolume is smaller than expected.

2.6.1 Integral Relationships

The fraction of the exit stream that has resided in the reactor for a period of

time shorter than a given value / is equal to the sum over all times less than t

QiE($)At, orexpressed continuously,

r

JE(t)dt =

Analogously,

jE(t)dt -

Fraction of effluent, which

has been in reactor for less

than time t

Fraction of effluent, which

has been in reactor for

longer than time t

=F(t) (14)

l-F(t) 05)

Because t appears in the integration limits of these two expressions, equation

(14) and (15) are both functions of time. F(t) can be calculated at various

times t from area under the curve of E(t) versus tplot.


ideal

.es

Wf

Figure2,3 (a) RTD for near plug flow reactor; (b) RTD for near perfectly mixedCSTR; (c) RTD for packed-bed reactorwith dead zones and channeling; (d) packed-bedreactor; (e) tank reactor with short-circuiting flow (bypass); (f) CSTRwithdead

zone.

19


2.6.2 Mean Residence Time

A parameter frequently used in analysis of ideal reactors is the space-time or

average residence time, r, which is defined as being equal to V/v. It can be

shown that no matter what RTD exists for a particular reactor, ideal or non-

ideal, this nominal holding time, r, is equal to the mean residence time, tm

As is the case with other variables described by its distribution functions, the

mean value of the variable is equal to the first moment of the RTD

function, £(/). Thus, the first moment is the mean residence time,--/- CO

MO* . K(0* V/rA,t =" =?«£(/)* =-2 =^' ' ' (16)m J V A1 co VTA, V /

\Eif)dt ° \c(t)dt ^CA''u 0

It is reported that the mean residence time of the liquid decreased with

increase in liquid space velocity. However, the mean residence time was a

stronger function of space velocity for the upflow mode of operation. The

higher mean residence time in the upflow mode could definitely provide a

better utilisation of catalyst. At the same time, the liquid would also spend

undesired longer residence time when not in contact with the catalyst. As a

result, a number of undesirable thermal reactions would take place during this

period.

When a larger size of diluent was used, the mean residence time increased

with gas/liquid ratio for the upflow mode of operation. The increased gas flow

rate in the upflow mode perhaps induced circulatory motion of liquid inside

the catalystbed so that the liquid spent more time in the reactor.

20


2.6.3 Other Moments of the RTD

It is very common to compare RTDs by using their moments instead of trying

to compare their entire distributions.

The second moment commonly used is taken about the mean and is called the

variance, or squareof the standard deviation. It is defined by

a%=\{t-tjE{t)dt (17)

Alternatively,

CO

\{t~tjc(t)dt yit_tfcAta2 =a =L{:' t-)C>*'1 (18)

\c(t)dt ^ ' '0

The magnitude of this moment is an indication of the 'spread' of the

distribution as it passes the vessel exit and has units of(time)2; the greater the

value of this moment, the greater a distribution's spread.

It is particularly usefiil for matching experimental curves to one ofa family of

theoretical curves. Figure 2.4 illustrates these terms.

Figure 2.4 Variance for matching theoretical curves

The third moment is also taken about the mean and is related to the skewness.

The skewness is defined by

21


1 *s"=^\{t-tjEif)dt (19)° o

The magnitude of this moment measures the extent that a distribution is

skewed in one direction or another in reference to the mean.

Rigorously, for complete description of a distribution, all moments must be

determined. Practically, these three (tm, a2, s3) are usually sufficient for a

reasonable characterization of an RTD.

2.7 Axial Dispersion

Suppose an ideal pulse of tracer is introduced into the fluid entering a reactor. The

pulse spreadsas it passes through the vessel. Therefore, to characterize the spreading,

according to dispersion model (Figure 2.5), it is assumed a diffusion-like process

superimposed on plug flow. This is called dispersion or longitudinal dispersion to

distinguish it from molecular diffusion. The dispersion coefficient D (m2/s) represents

this spreading process. Thus,

- large D means rapid spreading ofthe tracer curve

- small D means slow spreading

- D = 0 means no spreading, hence plug flow

Also, is the dimensionless group characterizingthe spread in the whole vessel.

D or D/uL is evaluated by recordingthe shape of the tracer curve as it passes the exit

of the vessel. In particular, tm (mean time of passage, or when the curve passes by the

exit) and a2 (variance, or a measure of the spread of the curve) are measured.

22

Dissertation

puiseci tracer—.,

at time *~- 0 \\

Pulse InputiiVinput)

ECB 5034 Final Year Research Proiect

The pulse starts spreading and this can beijausfiG o-y rTi&nv iTiingS: v.Siocij.y prcjfiiSj

iy?"by!etit mixing, molecular diffusion, etc.

Symmetrica? and gau^isn3i fitly M'iSiaf'i

iyh^surernentrvsint

Figure 25 The spreading of tracer according to the dispersion model

These measures, tm and a2, which are earlier mentioned, are directly linked by

theory to D and D/uL.

Consider plug flow of a fluid, on top of which is superimposed some degree of back-

mixing, the magnitude of which is independent of position within the vessel. This

conditionimpliesthat there exist no stagnant pockets and no gross bypassing or short-

circuiting of fluid in the vessel This is called the dispersed plug flow model, or

simply the dispersion model. Figure 2.6 shows the conditions visualized- Note that

with varying intensities of turbulence or intermixing the predictions of this model

should range from plug flow at one extreme to mixed flow at the other. As a result,

the reactor volume for this model will lie between those calculated for plug and mixed

flow.

pfGI'uS

1 I*"

1 d{— *j

^actuations due to different How

and turbulent diffusion /

FiuglV:'*' Disperse! plug fioVJ

Figure 2,6 Representation of the dispersion (dispersed plug flow) model

23


Since the mixing process involves a shuffling or redistribution of material either by

slippage or eddies, and since this is repeated many, many times during the flow of

fluid through the vessel, these disturbances are considered to be statistical in nature,

somewhat as in molecular diffusion. For molecular diffusion in the x-direction, the

governing differential equation is given by Fick's law;

^L=D —dt ~ dx2

(20)

Where D, the coefficient of molecular diffusion, is a parameter which uniquely

characterizes the process. In an analogous manner, it can be considered that all the

contributions to intermixing of fluid flowing in the x-direction to be described by a

similar form of expression, or

^ =D^r (21)dt dx2

Where the parameter D, which is called the longitudinal or axial dispersion

coefficient, uniquely characterizes the degree of back-mixing during flow. The terms

longitudinal and axial are used because it is to distinguish mixing in the direction of

flow from mixing in the lateral or radial direction, which is not the primary concern.

These two quantities may be quite different in magnitude. For example, in streamline

flow of fluids through pipes, axial mixing is mainly due to fluid velocity gradients,

whereas radial mixing is due to molecular diffusion alone.

In dimensionless form where z = (ut +x)lL and6 - tjtm = tujL, the basic differential

equation representing this dispersion model becomes

dC (' D\

d9 uLVMW

8C dC (22)dz1 az

(—

parameter that measures the extent of axial dispersion. Thus,

-> 0 Negligible dispersion, hence plug flow

—> go Large dispersion, hence mixed flow

Where the dimensionless group called the vessel dispersion number, is the

D\

uL

24


The dispersion model usually represents quite satisfactory flow that deviates not too

greatly from plug flow, thus real packed bed and tubes (not long ones if flow is

streamline).

The bed Pecletnumber (henceforth only Peclet number) of liquid is the reciprocal of

the dispersion number,yULj

, i.e.

Pe=-4r (23)D/uL

Whichthe dispersion number is also definedby

And

fD} al

\uLj(24)

al U a2 (25)

Where a2 is the variance ofthe E curve.

Peclet number value increased with the increase in liquid space velocity indicating the

reduction of back-mixing with higher liquid flow rate. The use of fine size of diluent

also reduced the dependency of Peclet number on space velocity. Increasing trend in

Peclet numberwith increasing liquid velocity for the upflow mode of operationwhen

a larger size of particles was used.

The Peclet number was a very strong decreasing function of gas/liquid ratio for the

upflow modewhen a larger size ofdiluent was used. The gas velocityaffects the axial

dispersion coefficient for upflow operation,

2.8 Stagnant Zone Volume

The hydrodynamics in a reactor is an important factor that influences the efficiency of

a reactor. The existence of stagnant zone greatly reduces the efficiency and

25


performance of the reactor. RTD analysis provides a good indication on the presence

of stagnant zone as well as the flowpattern throughthe reactor.

The deviation of the ideal flow can be determined by obtaining a complete velocity

distribution profile measured through the reactor. However, the approach is rather

impractical. Thus, there is a need to study the age of distribution of fluid exiting the

reactor by the stimulus response technique. The concept of the technique is to

introduce a tracer at the inlet or some point within the reactor. Then, at some point

along the reactor or at an exit, the tracer is collected to measure the concentration

subsequent time interval. In order to illustrate the RTD of the actual flow, the

stimulus-response experiment can be conducted with an appropriate choice of tracer.

The packed bed reactor presumably behaves as a plug flow reactor. However,

deviation from the ideal plug flow can occur due to short-circuiting, channeling or an

existence of dead zone (Figure 2.7). Arrangement of packing and adequate

distribution of liquid can disrupt the ideal behaviorofplug flow due to the channeling

of liquid.

An ideal plug flow behavior in which the tracer should emerge in the exit until

Ti~Td at the same concentration of the entrance. The mean residence time, tm is

calculated from RTD analysis, previously mentioned.

26


sno-i nr.i, un*

U- * if

v ~*\

v_^r Vj

^'J J

M

<J> -..V

\

••&

">.

Packed hed —-\

Channel mg, especial !yserious mcoimtercurreni

j two-phase operations

,^J

r xlrefTie short-circuitingand bvcHSss

Figure 2.7Non-ideal flow patterns, which may exist in process equipment

The mean residence time can be determined by the equation:

t. =

YCA'<(26)

The mean residence time can also be defined as the reactor volume-volumetric flow

rate ratio:

T -Y-d Q

(27)

The stagnant zone volume can be estimated based on the ratio of actual, tm and

theoretical HRT, Td:

r t A1—2V =Vstagnant (28)

27


If the tracer peak emerged earlier than the predicted theoretical HRT, this meant that

the effective volume of the reactor is reduced due to a form of channeling in the

packing media, which will give low tJTdratio.

It is also observed that the peak of higher flow rate will appear first, which indicated

the phenomenon of channeling. Another deviation is the tailing effect of the tracer

toward longer time, which indicated recycling effect and tracer accumulation in the

reactor.

2.9 Late Mixing

In a 'perfectly mixed' of a reactor, the entering fluid is assumed to be distributed

immediately and evenly throughout the reacting mixture. This mixing is assumed to

take place even on the micro scale, and elements of different ages mix together

thoroughly to form a completely micromixed fluid. If fluid elements of different ages

do not mix at all, the elements remain segregated from each other and the fluid is

termed completely segregated or late mixing. The extremes ofcomplete micromixing

and complete segregationare the limits of the micromixing of a reacting mixture.

For a continuous-flow system, each exit stream corresponds to a specific residence

time in the reactor. Batches of molecules are removed from the reactor at different

locations along the reactor in such a manner so as to duplicate the RTD function, Eft).

The molecules removed near the entrance to the reactor correspond to those molecules

having short residence times in the reactor. Physically, this effluent would correspond

to the molecules that channel rapidly through the reactor. The farther the molecules

travel along the reactor before being removed, the longer their residence time. The

points at which the various groups or batches of molecules are removed correspond to

the RTD function for the reactor.

Because there is no molecular interchange between globules, each acts essentially as

its own batch reactor. The reaction time in any one of these tiny batch reactors is

equal to the time that the particular globule spends in the reaction environment. The

28

Dissertation ECB 5034Final Year ResearchProiect

distribution of residence times among the globules is given by the RTD of the

particular reactor.

2.10 Early Mixing

In a reactor with segregated fluid, mixing between particles of fluid does not occur

until the fluid leaves the reactor. The reactor exit is, of course, the latest possible point

that mixing can occur, and any effect of mixing is postponed until after all reaction

has taken place. As soon as the fluid enters the reactor, it is completely mixed

radically with the other fluid already in the reactor. The entering fluid is fed into the

reactor through the side entrances in such a mannerthat the RTD the reactor with side

entrance is identical to the RTD of the real reactor. In the reactor with side entrances,

mixing occurs at the earliest possible moment consistent with the RTD. Thus, the

effect of the mixing occurs as much as possible throughout the reactor, and this

situation is termed the condition of maximum mixedness or early mixing. Maximum

mixedness is consistent with a given residence time distribution.

29


CHAPTER 3

PROJECT OVERVIEW

3.1 Research Project Work

Start

- r

Problem statement and

research justification

1'

Background research

Literature review

- Internet (onlineJournal)

- Journals

- Reference books

'i

Experimental work 4 M

1 '

Calcu lations/Spreadsheet

1 '

Data Analysis

Yes ,

No

'

1 ' taConclusion

1 '

Report Findings

1 '

End

Figure 3.1 Research methodology

The research was started with the problem statement identification and justification.

Once confirmed, further background research was done. This was accompanied by a

series of literature review, which is a continuous and on-going process throughout the

30


research. The literature review was done from several resources, mainly the internet

and the Information and Resource Center, Universiti Teknologi PETRONAS. This

includes journals, reference booksandalsoonline services suchas onlinejournals.

Once all the theoretical and findings from previous researches were firmed, sets of

experiments were conducted. The results of the experiments were later analysed and

compare with the theory and findings from precious researches. If the results were not

satisfactory, the experiment was repeated. After analysed, the results were then

compiled and conclusions were made.

3.2 Experimental Set-up

The simplified schematic diagram for the experimental set-up used is shown in

Figure 3.2. This set-up consists of a packed bed reactor, a system for feeding

controlled amount of gas and liquid, tracer injection system and conductivity

measurement instrument (conductivity probe) for detecting the concentration of

tracer. The data obtained from the experiment is recorded through data Acquisition

System (DAS).

Prom compressedaii line

FT1

Packed

bed

reactor

\ / i

I '

/ \I \

FTZ

,-p, conductivity'•-~J probe

Waste

lank

T3

liquid flewmeter

liquidpump

Dosing Tank(Tracer)

Feed rank

T2

Tt

Figure 3.2 Process Diagram for RTD Studied in Tubular Reactor (BP 112)

31


The reactor is essentially a column madefrom borosilicate glass with columninternal

diameter of 82 mm and a height of 1500 mm. The reactor is packed with 8x8 mm

Raschig rings. Bothtop and bottom of the reactor is fitted withstainless steel capwith

fitted ports for differential pressure tapping.

In this study, air andwaterused as the gas and liquid phase due to the inexpensive and

non-hazardous nature of these two materials. Sodium chloride acts as the tracer

medium. The flow rate of air and water into the system is controlled using mass flow

meter and the exit concentration of the tracer medium was measured using a

conductivity probe. Table 3.1 summarise the range of gas and liquid flow rate

considered for the experiment.

Before starting the experiment, the bed was fully wetted with water. Then, the air and

water flow rate is slowly adjusted to the desired values. The outlet valve of water is

adjusted until water level is at the same height as the bed. The tracer was introduced

through the reactor as a pulse through the dosing pump. Extra care was taken in

deciding the length of time to introduce the tracer. The tracer injection period must

take place over a short period oftime in comparison with the residence time to ensure

negligible dispersion between the point of injection and entrance to the reactor.

Operating liquid hold-up of a system is defined as the ratio ofthe amount of liquid in

the reactor during operation to the total volume of the packed bed reactor. The

operating liquid hold-up is determined as follows. The bed was fully wetted and the

liquid and gas flow rate was then adjusted to the set value for the experiment. The

system was left to attain a steady state condition, usually takes about 30 minutes.

After reaching steady state, the gas and liquid flow is stopped simultaneously. The

total free liquid in the reactor was drained in a liquid collector and measured.

32


Table 3.1 Experiment gas and liquid flow rateswith specified gas/liquid ratio

Gas Flowrate

(LPM)

7.5

Liquid Flowrate(LPM)

Gas/Liquid Ratio(LPM/LPM)

0.05

15015.0 0.10

22.5 0.15

10.0 0.05

20020.0 0.10

30.0 0.15

12.5 0.05

25025.0 0.10

37.5 0.15

33


CHAPTER 4

RESULTS AND DISCUSSION

4.1 Effect of Gas/Liquid Ratio on Pressure Drop and Operating Liquid

Hold-up

4.1.1 Effect of Gas/Liquid Ratio on Pressure Drop

Pressure drop analysis across the reactor is done by investigating the effect of

gas/liquid ratio. The pressure drop throughout the experiment is recorded and the

result is shown in Figure 4.1. For all constant liquid flow rates, the pressure drop

increases with increasing gas/liquid ratio.

Q.131 -

0.130 - A 0.130 i

0.129- / |

0.128 -/

j

_ 0.127 - / i.

re

•a

B~ 0.126 -go

/5 0.125 - *ai25 ^0.125 '.

in

£

°" 0.124-

0.123 •/

/ I

0.122 -/

/|

•=«™L=0.05LPM

0.121/ ™*»"L=0.10LPM

1 ' '

100 150 200

Gas/Uquid ratio (LPM/LPM)

250 300

Figure 4,1 Effect ofgas/liquid ratio on pressure drop at constant liquid flow rate of0.05 and 0.10 LPM

34


The results obtained in this experiment is in strong agreement with results obtained by

Varma et al. [4], who shows that the pressure drop increased with the gas and liquid

flow rates in all the regimes.

4.1.2 Effect of Gas/Liquid Ratio on Operating Liquid Hold-up

Liquid hold-up has a direct influence on the liquid side mass transfer, loading

behaviour and gas phase pressure gradients. Experimentally, operating liquid hold-up

is the portion of liquid that is drained out of a catalyst bed when both gas and liquid

are stopped after reaching steady state.

0.60-

0.50-

"^»aj4

X. 0.40 -a.

oX

•a

o- 0.30 -3

0.20 -

™^™L=0.05LPM

0.10-

™35r"L=0.10LPM

0.00-

100 150 200 250 300

GasrtJquid ratio (LPM/LPM)

Figure 4.2 Effect ofgas/liquid ratio on operating liquid hold-up at constant liquidflow rates of 0.05 and 0.10 LPM

Figure 4.2 shows that the operating liquid hold-up decreases with increasing

gas/liquid ratio. This is similar to the findings by Chander et al. [5]. Also, as the liquid

flow rates increase, the operating liquid holdup decreases. Thus, higher liquid flow

rates could increase the reaction rate in the three phase of packed bed reactor.

35


Study by Chander et al. [5] showed that for the downflow mode of operation, the

liquid hold-up was independent ofgas flow rate for large and small sizes ofdiluent.

4.2 Effect of Gas/Liquid Ratio on Moments of RTD

The variation in concentration data obtained from the experiments is shown as an E

curve. The values of Peclet number and mean residence time of the liquid were

calculated and analysed from the RTD data obtained from the experiment.

For this experiment, RTD experiment with pulse input is used. An amount of tracer,

NaCl, is injected in one shot into the feedstream entering the reactor in as short time

as possible. The outlet conductivity is then measured as a function of time. The

effluent concentration-time curve is referred as C curve in RTD analysis. However,

the consideration is more to the E curve and the three moments ofRTD.

4.2.1 Effect of Gas/Liquid ratio on E curve

The E curve is the distribution needed to account for non ideal flow. All three E

curves (Figure 4,3, Figure 4.4 and Figure 4.5) showed that at any gas/liquid ratio,

with low liquid flow rate, which in this case is 0.05 LPM, the E curve exhibited

deviation from ideal plug flow reactor and approached mixed flow behaviour.

However, as the liquid flow rates are increased, the E curves approach the behaviour

of a plug flow. From these three graphs, it is clearly showed that the concentration

profile of tracer is unaffected by the gas/liquid ratio but is a strong function of liquid

flow rates of the system.

36

Dissertation

0.035 T - - -•

0.020 -

0.015 -

0.005- ^ +


100 150

Time, minutes

»-G=7.5LPM.L=0.05LPM

*;-G=15.0LPM, L=0.10LPM

250


There are some fluctuations at concentration-time tail in Figure 4.3 and Figure 4.4

due to equipment failure in mass flow meter. The mass flow meter needs maintenance

for better results during the experiment in the future.

0.025 -,

&**&

0.020 -i

0.015 -

O.010 •

0.005 -

0.000

100

Tims, minutes

-G=10.0LPM, L=0.05LPM

-G=20.0LPM, L=0.10LPM


37

200

Dissertation

0.035 i ™

1.030 -

0.025 -

0.020 -

-0.015-

1.010-

0.005

0.000 3SS

0


150 200

Time(min)

*-G=12.5LPM, L=0.05LPM

5-G=37.5LPM, L=0.15LPM


350

4.2.2 Effect of Gas/Liquid Ratio on Mean Residence Time

0.90 -r

0.10 -

0.00 w

Time (min)

-s-G=12.5LPM. L=0.05LPM

-[£-G=37.5LPM. L=0.15LPM

Figure4.6 Mean residence time for gas/liquid ratio of 250

38

350


100 -,

90-Jg 83.21

!

80 -

69.46 JT

i

70 -

| 60-

mo

£ 50 -•o

o

|

i 40- ^^^- 38.81£ i

30-

20

Jf2f82i

•"*=-L=0.05LPM

10-^™L-0.10LPM I

100 150 200 250 300

Gas/Liquid ratio (LPM/LPM)

Figure 4.7 Effectof gas/liquid ratio on meanresidencetime of liquidat constantliquid flow rates of0.05 and 0.10 LPM

The nominal holding time, x, or the space time describes the time the material spends

in an ideal reactor. For a closed non ideal system with no dispersion, this should

equals to the mean residence time, xm, ofthe reactor.

Example of mean residence time curve is shown in Figure 4.6 for gas/liquid ratio of

250. Experimentally, it is observed that there is a considerable increase in mean

residence time as the gas/liquid ratio is increased, illustrated in Figure 4.7.

Furthermore, the mean residences time is increased with decreasing variation of liquid

flow rates. These results are also agreed with Chander et al. [5], which proved that the

mean residence time is increased with gas/liquid ratio for both, downflow and upflow,

mode ofoperation.

The analysis shows that the mean residence time for this reaction is high.

Furthermore, the peakof the E curves (Figure 4,3, Figure 4.4 andFigure4.5) occurs

earlier than the mean residence time. This is an indication of the possibility of

stagnant zone in the reactor. This could be due to excessive liquid holdup inside the

catalyst bed. According to Chander et al. [5], the higher mean residence time would

39


provide a better utilization ofcatalyst orthe liquid would also spend undesired longer

residence time when not in contact with the catalyst. In this case, the emerging peak

ofE curves, which is earlier than the mean residence time, suggests that there would

be insufficient contact with catalyst and also, excessive thermal reaction would not

likely to occur.

4.2.3 Effect of Gas/Liquid Ratio on Variance

The next most important descriptive quantity isthe spread ofthe distribution which is

representedby variance.

40.00 -i

35.00 -

30.00 -

25.00 -

20.00 -

15.00 -

10.00-

5.00-

150 200

Time(min)

—*-G=12.5LPM, L=0.05LPM

-3-G=37.5LPM, L=0.15LPM

Figure 4.8 Variance curve for gas/liquid ratio of 250

40


Jg 3167.59

j

3000 •

// i

2500 • //

/ i

"«= 2000 •I

tuoc

/ ]

/SJ582.13 /

ffl 1500 -

"^""^rfl100.94•mooJ

"*."»L=0.05LPM i500- _^-"& 520.65

^-^^^^^ —&—L=0.1QLPM |

i

2Z*ff&M

100 150 200 250

Gas/Liquid ratio (LPMJLPM)

300

Figure 4.9 Effect of gas/liquid ratio on variance at constant flow rates of0.05 and0.10 LPM

Example of variance curve is illustrated in Figure 4.8, which is for gas/liquid ratio of

250. Experimentally, from Figure 4.9, the variance increases as the gas/liquid ratio

increases. The variance is also increased with decreasing variation of liquid flow

rates. These results are also consistent with the E curve in Figure 4.3, Figure 4.4 and

Figure 4.5; the E curve for highest constant liquid flow rate has the highest peak

among the other constant liquid flow rates for all gas/liquid ratio, which indicates

smaller variance, and vice versa. Thus, it is proved that the smaller the variance, the

smaller the distribution's spread.

In order to achieve plug flow characteristic, smaller variance is required. Thus, from

this experiment, it is proved that operating packed bed reactor at high gas/liquid ratio

with lower gas and liquid flow rate, would result smaller variance, which is

approaching plug flow behavior.

41

Dissertation

4.2.4 Effect of Gas/Liquid Ratio on Skewness

2500-

1500 -

-1500 -

-2500

Tlme(min)

ECB 5034 Final Year Research Project

-G=12.5LPM, L^0.05LPM

-:-G=37.5LPM. L=0.15LPM

300 360

Figure 4.10 Skewness curve for gas/liquid ratio of 250

600- - -- —

i

500 - m 496.17

i

!

400-

/

//

I

c

i.™«a 300 •in<ucS0)

U)

200-

//

//

//

I

j

j

100 -

.A 149.03

^^ •"*S*"*L=0.05LPM

^^ ~*~L=0.10LPM

A*5a57!

100 150 200 250

Gas/Liquid ratio (LPM'LPM)

300

Figure 4.11 Effect of gas/liquid ratio on skewness at constant liquid flow rates of0.05andO.lOLPM

42


The skewness curve for gas/liquid ratio of 250 is shown in Figure 4.10. Analysis of

experiments have resulted that the skewness increases as the gas/liquid ratio increases,

refer to Figure4.11, and the skewness also increases as variation of liquid flow rates

decreases. Again, these results are reflected by the E curve in Figure 4,3, Figure 4.4

and Figure 4.5; the E curve for high constant liquid flow rate is not skewed far from

the reference of mean compared to other constant liquid flow rates for all gas/liquid

ratio, which indicates low skewness, and vice versa. Thus, it is proved that the lower

the skewness, the less skewed the distribution is, from its mean.

In order to achieve plug flow behavior, it is desirable to have smaller value of

skewness, which also meant that the distribution is skewed less from the reference of

the mean. Unlike E curve for lower flow rates which has high value of skewness and

skewed more towards the left, it is preferred to operate packed bed reactor at higher

gas/liquid ratio with lower gas and liquid flow rates, which will result lesser skewness

and will approach the plug flow behavior.

43


4.3 Effect of Gas/Liquid Ratio on Axial Dispersion and Stagnant Zone

Volume by RTD Analysis

4.3.1 Effect ofGas/Liquid Ratio on Axial Dispersion

140000 -•- —- - -----

„ _.

,

120000-

100000 -

11117333.22

//

//

/

i

i

•g 80000 -=

*oat

£ 60000 - i$£8605.01

A 77143.70 /

/ // /

j

i

!

^-^-Z / i

40000 - ^^^840730.67i

l

20000 -j!f21988.67

-S™-L=0.05LPM

™*»-L=0.10LPM

I

i

100 150 200 250

GasfLiquid ratio (LPMfLPM)

300

Figure 4.12 Effect ofgas/liquid on liquid Peclet number at constant liquid flow ratesof 0.05 and 0.10 LPM

Axial dispersion of a system gives an indication to the degree of mixing inside a

reactor and is represented by the Peclet number. Lower value of Peclet number

indicates higher degree of dispersion in the system.

Experimentally, the Peclet number relatively increases with increasing gas/liquid

ratio, as well as decreasing variation of liquid flow rates, as illustrated in Figure 4.12.

However, these results do not agree with studies done by Chander et al. [5], who

reported that the Peclet number value increased with the increase in LHSV for modes,

downflow:and upflow, of operation indicating the reduction of back-mixing with

higher liquid flow rate.

44


Besides, Figure 4.13 shows that Peclet number is decreasing as gas/liquid ratio

increasing. This result was obtained from previous student who did the same research

for counter-current flow. Further study is currently on-going to investigate this

difference.

Peclet Number versus GIL ratio

180,000.00

160.000.00 ^^^

140,000.00

120.000.00- *

Q.

t^

E 100.000.00 -

z

y = -262.45K* 199544

1

S. 80.000.00 -•D

71a;

60.000.00 J

40.000.00 -

20,000.00 -

0.00

c 50 100 150 200

Gas/Llquld rarlo (L/L)

250 300 350

Figure 4.13 Effect of gas/liquid on liquid Peclet number from other research

45


4.3.2 Effect of Gas/Liquid Ratio on Stagnant Zone Volume

3.5 -

3.0- 1^.03

.^-^•7°2.5-

m^A^ <\ \ i

_ 2.0--j 1ft. 1.93 V

| \V

> 1.5-1^1.40

1.0-

•H9»-L=G.05LPM j0.5 -

-*™L=0.10LPM

1C 0 150 200

Gas/Liquid ratio (LPM/LPM)

250 300

Figure 4.14 Effect ofgas/liquid ratio on stagnant zone volume at constant liquid flowrates of 0.05 and 0.10 LPM

The existence of stagnant zone inside a reactor is undesirable as it reduces the active

volume of the reactor, hence reducing the efFiciency and the overall performance of

the reactor. The variation in stagnant zone volume at different liquid flow rates and

gas/liquid ratio is shown in Figure 4.14. It is observed that operating at higher

gas/liquid ratio or higher liquid flow rate can reduce the stagnant zone volume. This

result is in agreement with Sata et.al [7].

At low gas/liquid ratio with low liquid flow rate, the mean residence time is lower

than the predicted theoretical HRT, which means that the effective volume of the

reactor is reduced due to a form of channeling in the packing media, which will give

low tm lTd ratio. Another deviation is the tailing effect of the tracer towards longer

time, which indicated recycling effect and tracer accumulation in the reactor.

46


CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

5.1 Conclusions

Pressure drop, AP can be reduced with decreasing gas/liquid ratio. Thus, gas and

liquid flow rates should be maintain low as to reduce pressure drop across the reactor.

The theories introduced in this research are considered parallel with the experimental

results.

Operating liquid hold-up, H0, is decreasing with increasing gas/liquid ratio in

experiment. Also, the liquid hold-up is decreased with increasing liquid flow rates.

Thus, for a desirable process, liquid hold-up must be minimized, which can be

achieved at high gas/liquid ratio with high gas and liquid flow rates.

Mean residence time, tm is increased with increasing gas/liquid ratio. This can be

explained by the induced circulatory motion of liquid with higher flow rates. This is

desirable for better utilization of catalyst. However, too high mean residence time

would also result undesirable thermal reaction. Mean residence time is also increased

with decreasing variation of liquid flow rates. Thus, the desirable operation can be

done at high gas/liquid ratio with lower gas and liquid flow rates.

Variance, a2 is decreased if the gas/liquid ratio is decreased. Also, the variance is

decreased with increasing variation of liquid flow rates. Variance reflects the spread

of distribution. The more the distribution spread, the higher the value of variance,

which also results more towards mixed flow behavior. Small variance is desired for a

Fixed bed reactor to behave more towards plug flow, which can be achieved at low

gas/liquid ratio, but with higher gas and liquid flow rates. This is also reflected in the

E curves ofall gas/liquid ratios.

47


Skewness, s3 measured the extent that the distribution is skewed in one direction or

another in reference to its mean. From the experiment, skewness can be decreased as

the gas/liquid ratio is decreased and with increasing variation of liquid flow rates.

Skewness is undesirable because the higher the value of skewness, the further the

distribution is skewed from the mean, which will also deviate from plug flow

characteristics. Thus, lower value of skewness is preferred to operate the packed bed

reactor towards plug flow behavior. This can be done at lower gas/liquid ratio, with

higher gas and liquid flow rates.

Axial dispersion of liquid, expressed by Peclet number, Pe, is increases with

increasing gas/liquid ratio as well as decreasing variation of liquid flow rates. This

condition is contradicted with the theory. Further study is on going to find out the

difference between experimental and theoretical.

Stagnant zone volume, Vstagnant, canbe reduced with higher gas/liquid ratio, as well as

higher liquid flow rates. This is mainly because the mean residence time is lower than

the predicted HRT. Thus, the effect of non ideal reactor, which is caused by

channeling, dead zones, or short-circuiting, can be reduced at higher gas/liquid ratio

flow rates with higher gas and liquid flow rates.

The information obtained from this residence time distribution study may therefore

may help in improving the performance of a reactor as well as finding an optimum

operating condition for a particular process.

48


5.2 Recommendations

Recommendations outlined here is based on studies that can be done or extended for

future development of RTD analysis, or rather the investigation of hydrodynamic

characteristics of fixed bed reactor or packed bed reactor.

5.2.1 Effect of Different Tracer Concentration

The RTD is determined experimentally by injecting an inert chemical, molecule, or

atom, called a tracer, into the reactor at some time t - 0 and then measuring the tracer

concentration, C, in the effluent stream as a function of time. In addition to being a

non-reactive species that is easily detectable, the tracer should have physical

properties similar to those of the reacting mixture and be completely soluble in the

mixture. The latter requirements are needed so that the behavior of tracer will

honestly reflect that of the material flowing through the reactor. Coloured and

radioactive materials are the two most common types of tracers.

So that, hydrodynamic characteristic can be affected by type of tracer. For further

investigation in RTD analysis, it is necessary if the research could include the

different types of tracer or different tracer concentration, used in the experiment, to

observe and analyse the hydrodynamics characteristic in packed bed or fixed bed

reactor.

5.2.2 Effect of Diluent Size

When the catalyst was loaded with smaller size of diluent, the values of mean

residence time, Peclet number and liquid holdup is expected to increase for the

downflow mode. As a result of this, the hydrodynamics behavior for both upflow and

downflow modes ofoperation can be improved.

Since the project study did not discuss the effect of diluent size, future study can be

made on investigating the change in the behavior of the upflow mode on using a

smaller size ofparticle as diluent in the catalyst bed. The use of smaller size ofdiluent

49


can increase the value of Peclet number and moderate the excessive liquid holdup,

and thus eliminated the limitations of the upflow mode of operation. The differences

in the nature of E curves for the two modes of operation under similar operating

conditions of liquid and gas velocities can also be eliminated for the smaller size of

diluent. The values of mean residence time, Peclet number and liquid holdup are

predicted nearly the same for the two modes of operation. Thus, the use of a smaller

size of diluent could remove the drawbacks of both upflow fixed bed and trickle bed

reactors, which will provide suitable tools for generating reliable data for scale-up and

scale-down activities.

The use of a smaller size of diluent can also decreased the porosity of the bed, which

in turn reduced the excessive mean residence time of liquid in the upflow mode of

operation. This could help in the reduction ofundesirable non-catalytic reaction in the

upflow mode.

50


REFERENCES

1. niuta, I.; Bildea, S.C; Iliuta, M.C.; Larachi, F.: Analysis ofTrickle Bed and

Packed Bubble Column Bioreactors for Combined Carbon Oxidation and

Nitrification, Braz. J. Chem. Eng. Vol. 19 no. 1 Sao Paulo Jan./Mar. 2002

2. Ergun, S.: Fluid Flow Through Packed Columns, Chemical Engineering

Program, 48, 89-94 (1952)

3. Turpin, J.L.; Huntington, R.L.: Prediction ofpressure dropfor twophase

two component concurrentflow in packed beds, AIChE J., 13 (1967) 1196

4. Varma, Y.B.G.; Khan, A.A.; Khan, A.: Flow regime identification and

pressure drop in co-current gas-liquid upflow through packed beds,

Bioprocess Engineering 16 (1997) 355-360 6 Springer-Verlag (1997)

5 Chander, A.; Kuiidu, A.; Bej, S.K.; Dalai, A.K.; Vohra, D.K.;

Hydrodynamic characteristics of cocurrent upflow and donwflow ofgas and

liquidin afixed bedreactoe, Fuel 80(2001) 1043 - 1053

6. Stiegel, G.J.; Shah, Y.T.: Industrial Engineering Chemical Process Design

Dev., 16(1), 37(1977)

7. Sata, S.A.; Mohamed, A.R; Kamaruddin, A.; Hydrodynamics experiment

in the packed bed reactor, 15th Symposium ofMalaysian Chemical EngineersSOMChE2001 (A3-2)

8. Fogler, H. Scott; Elements of Chemical Reaction Engineering, 2nd Edition,

Prentice Hall International Series, New Jersey, 711 - 725 (1992).

9. Perry's Chemical Engineers' Handbook, 7th Edition, Mac Graw Hill.

10 Nikolic, L.B.; Nikolic, V.D.; Veljkovic, V.B.; Lazic, MX.; Skala. D.U.;

Axial dispersion of the liquidphase in a three-phase Karr reciprocatingplate

column, J. Serb. Chem. Soc. 69 (7) 581 - 599 (2004)

11. Serbula, M.S.; Stankovic.; Hydrodynamic characteristics ofa two-phase gas

liquidflow upward through a fixed bed ofsphericalparticles, J. Serb. Chem.

Soc. 66(1)53-64(2001)

12. Levenspiel, O.: Chemical Reaction Engineering, 2nd Edition, John Wiley &

Sons, 253-330(1972).

51


APPENDICES

APPENDIX A RTD Analysis for Gas/Liquid Ratio of 150

3000-

2000-

.2 1500

oi

0 100 150

Time, minutes

-*-G=7.5LPM. L=0.05LPM

-s-G=15.0LPM. L=0.10LPM

FigureA-1 Conductivity against time for gas/liquid ratio of 150

0.900-1--—

0.700 -

EP 0.500 -

S 0.400-

0.300-

0.200 -

0.000 ibsss

100 150

Time, minutes

Figure A-2 Mean residence time curve for gas/liquid ratio of 150

52

250

250

Dissertation

20.00 -

"~ 15.00 -

ECB 5034 Final Year Research Project

100 150

Time, minutes

-G=7.5LPM. L=0.05LPM

-G=15.QLPM,L=0.10LPM

Figure A-3 Variance curve for gas/liquid ratio of 150

2S0

2000 -i

1000-

-500-

-1500-1

Time, minutes

Figure A-4 Skewness curve for gas/liquid ratio of 150

53


APPENDIX B RTD Analysis for Gas/Liquid Ratio of 200

3000!

2500-

0.700 -

0.600-

0.000 *

Time, minutes

-+-G=10.0LPM, L=0.05LPM

-ra-G=20.0LPM, l_=0.10LPM

150

FigureB-1 Conductivity against time for gas/liquid ratio of 200

100

Time, minutes

-#--G=10.0LPM, L=0.05LPM

-s-G=20.0LPM. L=0.10LPM

150

FigureB~2 Mean residence time curve for gas/liquid ratio of 200

54

200

200

Dissertation

25.00 v

20.00-

15 00-

10.00 -

100

Time, minutes


-«^-G=10.0LPM. L=0.Q5LPM

-H-G=20.0LPM. L=0.10LPM

Figure B-3 Variance curve for gas/liquid ratio of 200

1500 -

~*~G=10.QLPM. L=0.05LPM i

-S!-G=20.0LPM, I=0.10LPMt

1000 -

500 -

/r

rvi

X*

LU

3sgm^s^g

i

i

0 i

t

-500 •\ /

100 150 2C0

V,

-1000 •-~ — ~ —• ----

Time, minutes

• - •••-••

Figure B-4 Skewnesscurve for gas/liquid ratio of 200

55


APPENDIX C RTD Analysis for Gas/Liquid Ratio of 250

2500 n- -

2000 -

« 1500-

O 1000 -

Tirne{min}

Figure C-l Conductivity against time for gas/liquid ratio of 250

56

investigations of hydrodynamics of fixed bed reactor

Documents